Interview Shepherd is a web application that helps users practice job interviews with an AI-powered interviewer. The platform simulates realistic interview scenarios, provides instant feedback, and supports multiple job roles and languages through customizable job post to be used for interviews.
- AI-driven interviewer simulates real interview flow and adapts to user responses
- Users can create and manage job scenarios
- Resume upload for tailored interview questions
- Real-time chat interface with AI
- AI-evalutaion with scoring and suggestions
- User authentication/authorization and account management
- Frontend: HTML, CSS (Tailwind), JavaScript
- Backend: Node.js, Express, MongoDB
- AI Provider: Google Gemini 2.5 Flash
- Chat: SSE (Server Side Event)
- Clone the repository.
- Set up environment variables in
backend/.env(see.env.template) and config infrontend/config.js. - Install dependencies in
backendfolders. - Start the backend server:
cd backend npm install node server.js - Serve the frontend (e.g., with Live Server or any static server).
The project is a final project in Intro to CEDT course, part of Computer Engineering and Digital Technology major. The group of 6 random students must contribute to the project in concept of LLM in Web Application within two weeks (along with serveral mid-term/final exams). The project must
- Not use any library except for ExpressJS and database-related library.
- Be a Single Page Application with vanila JavaScript.
- Be responsive for all platform.
- Contain full CRUD operations.
Note
Despite the packed schedule and limited experience, we were able to deliver this in three days, thankfully. I appreciate everyone's determination throughout the project. It was a very tough task for team who had only four three-hour web development classes, and the project period overlapped with actual web development class. Nevertheless, It was so much fun working with you all. — @14two-77